Transceivers are tuned to an output impedance that is designed to match the antenna to which the transceiver is operating. Traditionally, in the telecommunications environment, that output impedance is 50 ohms. If the output impedance of the transceiver and the input impedance of the antenna perfectly match, the output power from the transceiver will be fully carried to the antenna. On the other hand, if, as is more usually the case, the output impedance of the transceiver is not perfectly matched to the antenna, a portion of the output power is reflected back into the transceiver, rather than being carried to the antenna for transmission.
The reflection of power due to an imbalance between the transceiver load and the antenna is measured by a so-called voltage standing wave ratio. In quantitative terms, the voltage standing wave ratio for a transceiver at any given point in time is equal to the reflected power from the antenna divided by the output power from the transceiver. As shown in FIG. 1, for example, the transmitter/receiver 10 has a characteristic output impedance of 50 hms. In an ideal scenario, the impedance Z.sub.o of the antenna 12 would also exhibit the same 50 ohm characteristic at its input, thus matching the impedance of the transceiver. When this match situation occurs, the reflected power E.sub.r is zero and the voltage standing wave ratio (E.sub.r /E.sub.f) is zero as well, meaning that all of the transmitter power is delivered to the antenna for transmission.
But, when the antenna impedance varies, the impedance match condition vanishes and some of the transmitted power is reflected back to the transmitter. In other words, as the antenna impedance Z.sub.o moves from 50 ohms, the reflected power E.sub.r is greater than zero, and the voltage standing wave ratio is greater than zero. These changes in the antenna impedance Z.sub.o occur, for example, when people or other objects move nearer to or farther from the antenna 12, thus changing the antenna load characteristics. Such changes in the antenna load are quite common in mobile radio transmitters, which travel in and out of scenes that may put them closer to or farther from different types of human and non-human objects.
In fact, some mobile radio transceivers can exhibit dramatic swings in their instantaneous voltage standing wave ratio, caused mainly due to the presence (or lack of presence) of human and non-human objects in the vicinity of the transceiver antenna. The power amplifier in the transceiver is particularly sensitive to such swings, and thus must be designed to accommodate them. A problem exists, however, when high variation in the voltage standing wave ratio disturbs the linear operation of the power amplifier. In such a situation, cartesian feedback can be employed to try to maintain that linear operation, but even then, wide swings in voltage standing wave ratio can adversely effect the transceiver operation.
Portable transceivers have, in the past, tried to contend with the detuning that occurs at the antenna when the transceiver approaches and retreats from objects. Ordinarily, the solution has been to design power amplifiers in the transceiver to withstand a worst-case scenario. Sometimes, for example, where no amplitude variation in the RF carrier exists, the problem can be solved by designing a stable RF power amplifier at around 8:1 voltage standing wave ratio. But, most modern modulation techniques involve amplitude variation in the RF carrier (such as DQ Phase Shift Key Modulation, for example), which can cause the voltage standing wave variations to adversely affect the quality of the transmitted signal, even when the amplifier is designed at the 8:1 ratio. These problems are then made even worse when class AB or class C amplifiers are employed.
A Cartesian feedback linearization technique, discussed above, can settle some of the adverse quality problems associated with voltage standing wave variation. But, the effects of variation in voltage standing wave ratios can cause the Cartesian feedback loop to stretch beyond its ability to compensate. Indeed, maintaining stability in the cartesian feedback loop is vital to its good operation, and wide swings in voltage standing wave ratio can cause this loop to lapse into instability. Further complicating this problem is the difficulty that exists in detecting (and thus preventing) large numbers of production line products that may ultimately exhibit this instability problem.
Prior devices have attempted to ameliorate this problem using a variety of unsatisfactory techniques. For example, prior devices have employed an RF isolator (typically ferrite) ahead of the antenna to shield the power amplifier in the transceiver from any antenna load variation. Unfortunately, isolators are costly, operate in a narrow band, and are large (especially when employed at 150 MHz or higher).
Other prior devices have placed a fixed load between the power amplifier and the antenna in order to limit the voltage standing wave ratio. This technique effectively controls voltage reflection, but does so at a cost of output power. In fact, a fixed load will drain output power even when the antenna load perfectly matches the transceiver output load, i.e., when the reflected power is zero and no voltage standing wave ratio control is even needed.
A still further method employed by prior devices involves controlling the impedance of the antenna dynamically. In these devices, a tunable matching circuit between the antenna load and the output of the amplifier adjusts the load seen by the transceiver to the approximate 50 ohm level anytime the antenna load is dramatically altered by its surroundings. This re-matching technique can be effective, but it is a complicated method involving impedance detectors at the antenna and matching circuits that must be easily tunable in high voltage conditions like those exhibited at the output of the RF power amplifier.